CircleCI build openturns 16607

openturns/master/_downloads/22f8ab9d416a8b8e143fbc52b064158d/plot_nais.ipynb

@@ -165,6 +165,24 @@
         "quantileLevel = 0.1\nalgo = ot.NAIS(myEvent, quantileLevel)"
       ]
     },
+    {
+      "cell_type": "markdown",
+      "metadata": {},
+      "source": [
+        "In order to get all the inputs and outputs that realize the event, you have to mention it now:\n\n"
+      ]
+    },
+    {
+      "cell_type": "code",
+      "execution_count": null,
+      "metadata": {
+        "collapsed": false
+      },
+      "outputs": [],
+      "source": [
+        "algo.setKeepSample(True)"
+      ]
+    },
     {
       "cell_type": "markdown",
       "metadata": {},
@@ -223,7 +241,7 @@
       "cell_type": "markdown",
       "metadata": {},
       "source": [
-        "## Draw the NAIS samples used by the algorithm\n\nThe following manipulations are possible only if you have created a :class:`~openturns.MemoizeFunction` that enables to store all the inputs and outputs of the function $g$.\n\nGet all the inputs and outputs where $g$ were evaluated:\n\n"
+        "You can also get the successive thresholds used by the algorithm:\n\n"
       ]
     },
     {
@@ -234,14 +252,21 @@
       },
       "outputs": [],
       "source": [
-        "inputNAIS = g.getInputHistory()\noutputNAIS = g.getOutputHistory()\nnTotal = inputNAIS.getSize()\nprint(\"Number of evaluations of g = \", nTotal)"
+        "levels = algo.getThresholdPerStep()\nprint(\"Levels of g = \", levels)"
+      ]
+    },
+    {
+      "cell_type": "markdown",
+      "metadata": {},
+      "source": [
+        "## Draw the NAIS samples used by the algorithm\n\n"
       ]
     },
     {
       "cell_type": "markdown",
       "metadata": {},
       "source": [
-        "Within each step of the algorithm, a sample of size $N$ is created, where:\n\n"
+        "You can get the number $N_s$ of steps with:\n\n"
       ]
     },
     {
@@ -252,14 +277,14 @@
       },
       "outputs": [],
       "source": [
-        "N = algo.getMaximumOuterSampling() * algo.getBlockSize()\nprint(\"Size of each subset = \", N)"
+        "Ns = algo.getStepsNumber()\nprint(\"Number of steps= \", Ns)"
       ]
     },
     {
       "cell_type": "markdown",
       "metadata": {},
       "source": [
-        "You can get the number $N_s$ of steps with:\n\n"
+        "Get all the inputs where $g$ was evaluated at each step\n\n"
       ]
     },
     {
@@ -270,14 +295,14 @@
       },
       "outputs": [],
       "source": [
-        "Ns = int(nTotal / N)\nprint(\"Number of steps = \", Ns)"
+        "list_subSamples = list()\nfor step in range(Ns):\n    list_subSamples.append(algo.getInputSample(step))"
       ]
     },
     {
       "cell_type": "markdown",
       "metadata": {},
       "source": [
-        "Now, we can split the initial sample into NAIS samples of size $N_s$:\n\n"
+        "The following graph draws each NAIS sample and the frontier $g(x_1, x_2) = l_i$ where $l_i$ is the threshold at the step $i$:\n\n"
       ]
     },
     {
@@ -288,14 +313,14 @@
       },
       "outputs": [],
       "source": [
-        "listNAISSamples = list()\nlistOutputNAISSamples = list()\nfor i in range(Ns):\n    listNAISSamples.append(inputNAIS[i * N: i * N + N])\n    listOutputNAISSamples.append(outputNAIS[i * N: i * N + N])"
+        "graph = ot.Graph()\ngraph.setAxes(True)\ngraph.setGrid(True)\ngraph.setTitle(\"NAIS sampling: samples\")\ngraph.setXTitle(r\"$x_1$\")\ngraph.setYTitle(r\"$x_2$\")\ngraph.setLegendPosition(\"lower left\")"
       ]
     },
     {
       "cell_type": "markdown",
       "metadata": {},
       "source": [
-        "And get all the levels defining the intermediate and final thresholds given by the empirical quantiles of each NAIS output sample:\n\n"
+        "Add all the NAIS samples:\n\n"
       ]
     },
     {
@@ -306,14 +331,25 @@
       },
       "outputs": [],
       "source": [
-        "levels = []\nfor i in range(Ns - 1):\n    levels.append(listOutputNAISSamples[i].computeQuantile(quantileLevel)[0])\nlevels.append(threshold)"
+        "for i in range(Ns):\n    cloud = ot.Cloud(list_subSamples[i])\n    # cloud.setPointStyle(\"dot\")\n    graph.add(cloud)\ncol = graph.getColors()"
       ]
     },
     {
       "cell_type": "markdown",
       "metadata": {},
       "source": [
-        "The following graph draws each NAIS sample and the frontier $g(x_1, x_2) = l_i$ where $l_i$ is the threshold at the step $i$:\n\n"
+        "Add the frontiers $g(x_1, x_2) = l_i$ where $l_i$ is the threshold at the step $i$:\n\n"
+      ]
+    },
+    {
+      "cell_type": "code",
+      "execution_count": null,
+      "metadata": {
+        "collapsed": false
+      },
+      "outputs": [],
+      "source": [
+        "gIsoLines = g.draw([-5] * 2, [5] * 2, [128] * 2)\ndr = gIsoLines.getDrawable(0)\nfor i, lv in enumerate(levels):\n    dr.setLevels([lv])\n    dr.setLineStyle(\"solid\")\n    dr.setLegend(r\"$g(X) = $\" + str(round(lv, 2)))\n    dr.setLineWidth(3)\n    dr.setColor(col[i])\n    graph.add(dr)"
       ]
     },
     {
@@ -324,14 +360,14 @@
       },
       "outputs": [],
       "source": [
-        "graph = ot.Graph(\"NAIS samples\", \"x1\", \"x2\", True, \"lower left\")\ngraph.setGrid(True)"
+        "_ = View(graph)"
       ]
     },
     {
       "cell_type": "markdown",
       "metadata": {},
       "source": [
-        "Add all the NAIS samples:\n\n"
+        "## Draw the frontiers only\n\nThe following graph enables to understand the progression of the algorithm:\n\n"
       ]
     },
     {
@@ -342,14 +378,14 @@
       },
       "outputs": [],
       "source": [
-        "for i in range(Ns):\n    cloud = ot.Cloud(listNAISSamples[i])\n    graph.add(cloud)\ncol = ot.Drawable().BuildDefaultPalette(Ns)\ngraph.setColors(col)"
+        "graph = ot.Graph()\ngraph.setAxes(True)\ngraph.setGrid(True)\ndr = gIsoLines.getDrawable(0)\nfor i, lv in enumerate(levels):\n    dr.setLevels([lv])\n    dr.setLineStyle(\"solid\")\n    dr.setLegend(r\"$g(X) = $\" + str(round(lv, 2)))\n    dr.setLineWidth(3)\n    graph.add(dr)\n\ngraph.setColors(col)\ngraph.setLegendPosition(\"lower left\")\ngraph.setTitle(\"NAIS sampling: thresholds\")\ngraph.setXTitle(r\"$x_1$\")\ngraph.setYTitle(r\"$x_2$\")\n\n_ = View(graph)"
       ]
     },
     {
       "cell_type": "markdown",
       "metadata": {},
       "source": [
-        "Add the frontiers $g(x_1, x_2) = l_i$ where $l_i$ is the threshold at the step $i$:\n\n"
+        "## Get all the input and output points that realized the event\nThe following lines are possible only if you have mentioned that you wanted to keep samples with the method *algo.setKeepSample(True)*\n\n"
       ]
     },
     {
@@ -360,14 +396,14 @@
       },
       "outputs": [],
       "source": [
-        "gIsoLines = g.draw([-8] * 2, [8] * 2, [128] * 2)\ndr = gIsoLines.getDrawable(2)\nfor i, lv in enumerate(levels):\n    dr.setLevels([lv])\n    dr.setLineStyle(\"solid\")\n    dr.setLegend(r\"$g(X) = $\" + str(round(lv, 2)))\n    dr.setLineWidth(3)\n    dr.setColor(col[i])\n    graph.add(dr)\n\n# sphinx_gallery_thumbnail_number = 2\nview = View(graph)"
+        "select = ot.NAIS.EVENT1  # points that realize the event\nstep = Ns - 1  # get the working sample from last iteration\ninputEventSample = algo.getInputSample(step, select)\noutputEventSample = algo.getOutputSample(step, select)\nprint(\"Number of event realizations = \", inputEventSample.getSize())"
       ]
     },
     {
       "cell_type": "markdown",
       "metadata": {},
       "source": [
-        "## Draw the frontiers only\n\nThe following graph enables to understand the progression of the algorithm from the mean value of the initial distribution to the limit state function:\n\n"
+        "Draw them! They are all in the event space.\n\n"
       ]
     },
     {
@@ -378,7 +414,7 @@
       },
       "outputs": [],
       "source": [
-        "graph = ot.Graph(\"NAIS thresholds\", \"x1\", \"x2\", True, \"lower left\")\ngraph.setGrid(True)\ndr = gIsoLines.getDrawable(0)\nfor i, lv in enumerate(levels):\n    dr.setLevels([lv])\n    dr.setLineStyle(\"solid\")\n    dr.setLegend(r\"$g(X) = $\" + str(round(lv, 2)))\n    dr.setLineWidth(3)\n    graph.add(dr)\n\ngraph.setColors(col)\nview = View(graph)"
+        "graph = ot.Graph()\ngraph.setAxes(True)\ngraph.setGrid(True)\ncloud = ot.Cloud(inputEventSample)\ncloud.setPointStyle(\"bullet\")\ngraph.add(cloud)\ngIsoLines = g.draw([-5] * 2, [5] * 2, [1000] * 2)\ndr = gIsoLines.getDrawable(0)\ndr.setLevels([0.0])\ndr.setColor(\"red\")\ngraph.add(dr)\n_ = View(graph)\n\nView.ShowAll()"
       ]
     }
   ],

openturns/master/_downloads/f6ae28c1a237f6decc9385265d5edc53/plot_nais.py

@@ -108,6 +108,12 @@ def fourBranch(x):
 quantileLevel = 0.1
 algo = ot.NAIS(myEvent, quantileLevel)
 
+# %%
+# In order to get all the inputs and outputs that realize the event, you have to mention it now:
+
+# %%
+algo.setKeepSample(True)
+
 # %%
 # Now you can run the algorithm.
 
@@ -138,75 +144,55 @@ def fourBranch(x):
 )
 
 # %%
-# Draw the NAIS samples used by the algorithm
-# -------------------------------------------
-#
-# The following manipulations are possible only if you have created a :class:`~openturns.MemoizeFunction` that enables to store all the inputs and outputs of the function :math:`g`.
-#
-# Get all the inputs and outputs where :math:`g` were evaluated:
-
-# %%
-inputNAIS = g.getInputHistory()
-outputNAIS = g.getOutputHistory()
-nTotal = inputNAIS.getSize()
-print("Number of evaluations of g = ", nTotal)
+# You can also get the successive thresholds used by the algorithm:
 
 # %%
-# Within each step of the algorithm, a sample of size :math:`N` is created, where:
+levels = algo.getThresholdPerStep()
+print("Levels of g = ", levels)
 
 # %%
-N = algo.getMaximumOuterSampling() * algo.getBlockSize()
-print("Size of each subset = ", N)
+# Draw the NAIS samples used by the algorithm
+# ---------------------------------------------
 
 # %%
 # You can get the number :math:`N_s` of steps with:
+Ns = algo.getStepsNumber()
+print("Number of steps= ", Ns)
 
 # %%
-Ns = int(nTotal / N)
-print("Number of steps = ", Ns)
-
-# %%
-# Now, we can split the initial sample into NAIS samples of size :math:`N_s`:
-
-# %%
-listNAISSamples = list()
-listOutputNAISSamples = list()
-for i in range(Ns):
-    listNAISSamples.append(inputNAIS[i * N: i * N + N])
-    listOutputNAISSamples.append(outputNAIS[i * N: i * N + N])
-
-# %%
-# And get all the levels defining the intermediate and final thresholds given by the empirical quantiles of each NAIS output sample:
-
-# %%
-levels = []
-for i in range(Ns - 1):
-    levels.append(listOutputNAISSamples[i].computeQuantile(quantileLevel)[0])
-levels.append(threshold)
+# Get all the inputs where :math:`g` was evaluated at each step
+list_subSamples = list()
+for step in range(Ns):
+    list_subSamples.append(algo.getInputSample(step))
 
 # %%
 # The following graph draws each NAIS sample and the frontier :math:`g(x_1, x_2) = l_i` where :math:`l_i` is the threshold at the step :math:`i`:
 
 # %%
-graph = ot.Graph("NAIS samples", "x1", "x2", True, "lower left")
+graph = ot.Graph()
+graph.setAxes(True)
 graph.setGrid(True)
+graph.setTitle("NAIS sampling: samples")
+graph.setXTitle(r"$x_1$")
+graph.setYTitle(r"$x_2$")
+graph.setLegendPosition("lower left")
 
 # %%
 # Add all the NAIS samples:
 
 # %%
 for i in range(Ns):
-    cloud = ot.Cloud(listNAISSamples[i])
+    cloud = ot.Cloud(list_subSamples[i])
+    # cloud.setPointStyle("dot")
     graph.add(cloud)
-col = ot.Drawable().BuildDefaultPalette(Ns)
-graph.setColors(col)
+col = graph.getColors()
 
 # %%
 # Add the frontiers :math:`g(x_1, x_2) = l_i` where :math:`l_i` is the threshold at the step :math:`i`:
 
 # %%
-gIsoLines = g.draw([-8] * 2, [8] * 2, [128] * 2)
-dr = gIsoLines.getDrawable(2)
+gIsoLines = g.draw([-5] * 2, [5] * 2, [128] * 2)
+dr = gIsoLines.getDrawable(0)
 for i, lv in enumerate(levels):
     dr.setLevels([lv])
     dr.setLineStyle("solid")
@@ -215,17 +201,18 @@ def fourBranch(x):
     dr.setColor(col[i])
     graph.add(dr)
 
-# sphinx_gallery_thumbnail_number = 2
-view = View(graph)
+# %%
+_ = View(graph)
 
 # %%
 # Draw the frontiers only
 # -----------------------
 #
-# The following graph enables to understand the progression of the algorithm from the mean value of the initial distribution to the limit state function:
+# The following graph enables to understand the progression of the algorithm:
 
 # %%
-graph = ot.Graph("NAIS thresholds", "x1", "x2", True, "lower left")
+graph = ot.Graph()
+graph.setAxes(True)
 graph.setGrid(True)
 dr = gIsoLines.getDrawable(0)
 for i, lv in enumerate(levels):
@@ -236,4 +223,40 @@ def fourBranch(x):
     graph.add(dr)
 
 graph.setColors(col)
-view = View(graph)
+graph.setLegendPosition("lower left")
+graph.setTitle("NAIS sampling: thresholds")
+graph.setXTitle(r"$x_1$")
+graph.setYTitle(r"$x_2$")
+
+_ = View(graph)
+
+# %%
+# Get all the input and output points that realized the event
+# -----------------------------------------------------------
+# The following lines are possible only if you have mentioned that you wanted to keep samples with the method *algo.setKeepSample(True)*
+
+# %%
+select = ot.NAIS.EVENT1  # points that realize the event
+step = Ns - 1  # get the working sample from last iteration
+inputEventSample = algo.getInputSample(step, select)
+outputEventSample = algo.getOutputSample(step, select)
+print("Number of event realizations = ", inputEventSample.getSize())
+
+# %%
+# Draw them! They are all in the event space.
+
+# %%
+graph = ot.Graph()
+graph.setAxes(True)
+graph.setGrid(True)
+cloud = ot.Cloud(inputEventSample)
+cloud.setPointStyle("bullet")
+graph.add(cloud)
+gIsoLines = g.draw([-5] * 2, [5] * 2, [1000] * 2)
+dr = gIsoLines.getDrawable(0)
+dr.setLevels([0.0])
+dr.setColor("red")
+graph.add(dr)
+_ = View(graph)
+
+View.ShowAll()

openturns/master/_sources/auto_calibration/bayesian_calibration/plot_gibbs.rst.txt

@@ -283,7 +283,7 @@ Let us plot the posterior density.
 
 .. rst-class:: sphx-glr-timing
 
-   **Total running time of the script:** (0 minutes 14.208 seconds)
+   **Total running time of the script:** (0 minutes 12.689 seconds)
 
 
 .. _sphx_glr_download_auto_calibration_bayesian_calibration_plot_gibbs.py:

openturns/master/_sources/auto_calibration/bayesian_calibration/plot_rwmh_python_distribution.rst.txt

@@ -425,7 +425,7 @@ Plot posterior marginal plots only as prior cannot be drawn meaningfully.
 
 .. rst-class:: sphx-glr-timing
 
-   **Total running time of the script:** (0 minutes 5.709 seconds)
+   **Total running time of the script:** (0 minutes 5.875 seconds)
 
 
 .. _sphx_glr_download_auto_calibration_bayesian_calibration_plot_rwmh_python_distribution.py: